October 2023 arXiv papers — page 170

Haoxuan Liu, Vasu Singh, Michał Filipiuk, Siva Kumar Sastry Hari

Vision Transformers are being increasingly deployed in safety-critical applications that demand high reliability. It is crucial to ensure the correctness of their execution in spite of potential errors such as transient hardware errors. We propose a novel algorithm-based resilience framework called ALBERTA that allows us to perform end-to-end resilience anal

Fan Lei, Arlen Fan, Alan M. MacEachren, Ross Maciejewski

Visualization linting is a proven effective tool in assisting users to follow established visualization guidelines. Despite its success, visualization linting for choropleth maps, one of the most popular visualizations on the internet, has yet to be investigated. In this paper, we present GeoLinter, a linting framework for choropleth maps that assists in cre

Zhu Wang

Ontology matching (OM) entails the identification of semantic relationships between concepts within two or more knowledge graphs (KGs) and serves as a critical step in integrating KGs from various sources. Recent advancements in deep OM models have harnessed the power of transformer-based language models and the advantages of knowledge graph embedding. Never

Angel Israel Toledo Castro

One way to understand the deformation theory of a tensor category $M$ is through its Davydov-Yetter cohomology $H_{DY}^{\ast}(M)$ which in degree 3 and 4 is known to control respectively first order deformations of the associativity coherence of $M$ and their obstructions. \\ In this work we take the task of developing an analogous theory for the deformation

Harsh Chaudhari, Giorgio Severi, Alina Oprea, Jonathan Ullman

The integration of machine learning (ML) in numerous critical applications introduces a range of privacy concerns for individuals who provide their datasets for model training. One such privacy risk is Membership Inference (MI), in which an attacker seeks to determine whether a particular data sample was included in the training dataset of a model. Current s

Ken C. L. Wong, Hongzhi Wang, Tanveer Syeda-Mahmood

With the introduction of Transformers, different attention-based models have been proposed for image segmentation with promising results. Although self-attention allows capturing of long-range dependencies, it suffers from a quadratic complexity in the image size especially in 3D. To avoid the out-of-memory error during training, input size reduction is usua

Björn Müller, Thorsten Hohage, Damien Fournier, Laurent Gizon

In passive imaging, one attempts to reconstruct some coefficients in a wave equation from correlations of observed randomly excited solutions to this wave equation. Many methods proposed for this class of inverse problem so far are only qualitative, e.g., trying to identify the support of a perturbation. Major challenges are the increase in dimensionality wh

A. V. Yulin, E. S. Sedov, A. V. Kavokin, I. A. Shelykh

We predict the limit cycle solution for a ring-shape bosonic condensate of exciton-polaritons confined in an optically induced rotating trap. The limit cycle manifests itself with polarization oscillations on a characteristic timescale of tens of picoseconds. The effect arises due to the interplay between orbital motion and the polarization degree of freedom

Parthkumar Bhuvela, Hooman Taghavi, Adel Nasiri

An inverter is generally employed with MV LFT to connect to the grid in a grid-tied PV system. However, in some single-stage topologies, the LFTs are replaced by HFT combined with an unfolder inverter. Generally, these topologies have limited use at high-power MV grids due to high switching losses on the primary side. This study proposes an LLC resonant conv

Omar Zamzam, Haleh Akrami, Mahdi Soltanolkotabi, Richard Leahy

In this paper, we address the problem of learning a binary (positive vs. negative) classifier given Positive and Unlabeled data commonly referred to as PU learning. Although rudimentary techniques like clustering, out-of-distribution detection, or positive density estimation can be used to solve the problem in low-dimensional settings, their efficacy progres

Alexander J. Dittmann, Adam M. Dempsey, Hui Li

The accretion disks that fuel active galactic nuclei (AGN) may house numerous stars and compact objects, formed in situ or captured from nearby star clusters. Embedded neutron stars and black holes may form binaries and eventually merge, emitting gravitational waves detectable by LIGO/VIRGO. AGN disks are a particularly promising environment for the producti

Shahid Alam

A vital issue of file carving in digital forensics is type classification of file fragments when the filesystem metadata is missing. Over the past decades, there have been several efforts for developing methods to classify file fragments. In this research, a novel sifting approach, named SIFT (Sifting File Types), is proposed. SIFT outperforms the other stat

Petter Törnberg, Diliara Valeeva, Justus Uitermark, Christopher Bail

Social media is often criticized for amplifying toxic discourse and discouraging constructive conversations. But designing social media platforms to promote better conversations is inherently challenging. This paper asks whether simulating social media through a combination of Large Language Models (LLM) and Agent-Based Modeling can help researchers study ho

Reza Ghaiumy Anaraky, Byron Lowens, Yao Li, Kaileigh A. Byrne

Considering that prior research has found older users undergo a different privacy decision-making process compared to younger adults, more research is needed to inform the behavioral privacy disclosure effects of these strategies for different age groups. To address this gap, we used an existing dataset of an experiment with a photo-tagging Facebook applicat

Krishnadas Mohandas, Krzysztof Suchecki, Janusz A. Holyst

Heider's structural balance theory has proven invaluable in comprehending the dynamics of social groups characterized by both friendly and hostile relationships. Since people's relations are rarely single-faceted, we investigate Heider balance dynamics on a multiplex network, consisting of several copies of the same agent displaying correlated relations at d

Kamran Karimi, Amin Esfandiarpour, Stefanos Papanikolaou

Under plastic flow, multi-element high/medium-entropy alloys (HEAs/MEAs) commonly exhibit complex intermittent and collective dislocation dynamics owing to inherent lattice distortion and atomic-level chemical complexities. Using atomistic simulations, we report on an avalanche study of slowly-driven model face-centered cubic (fcc) NiCoCrFeMn and NiCoCr chem

Sneha Muppalla, Shan Jia, Siwei Lyu

Deepfakes are AI-generated media in which an image or video has been digitally modified. The advancements made in deepfake technology have led to privacy and security issues. Most deepfake detection techniques rely on the detection of a single modality. Existing methods for audio-visual detection do not always surpass that of the analysis based on single mod

Wei Gu, Leonardo C. Mihalcea, Eric Sharpe, Weihong Xu

In a recent paper, we stated conjectural presentations for the equivariant quantum K ring of partial flag varieties, motivated by physics considerations. In this companion paper, we analyze these presentations mathematically. We start by proving a Nakayama type result for quantum K theory: if the conjectured set of relations deforms a complete set of relatio

Qinglan Xia, Haotian Sun

In the Monge-Kantorovich transport problem, the transport cost is expressed in terms of transport maps or transport plans, which play crucial roles there. A variant of the Monge-Kantorovich problem is the ramified (branching) transport problem that models branching transport systems via transport paths. In this article, we showed that any cycle-free transpor

William J. Ugalde

A review of the characterization of principal bundles, through the different properties of the action of a group and its related canonical and translation maps, is presented. The work is divided in three stages: a topological group acting on a topological space, a discrete group acting on a smooth manifold, and a Lie group acting on a smooth manifold.

Atah Nuh Mih, Hung Cao, Asfia Kawnine, Monica Wachowicz

The use of edge devices together with cloud provides a collaborative relationship between both classes of devices where one complements the shortcomings of the other. Resource-constraint edge devices can benefit from the abundant computing power provided by servers by offloading computationally intensive tasks to the server. Meanwhile, edge devices can lever

Pedro Rizzo, Joel Torres Del Valle, Alexander Torres-Gomez

This article investigates the properties of Dedekind superrings, invertible supermodules and projective supermodules within the $\mathbb{Z}_2$-graded framework. Rather than treating these entities as specialized instances of general noncommutative ring theory, we develop them intrinsically within the category of supercommutative superrings. We examine the st

Tsung-Lin Tsou, Tsung-Han Wu, Winston H. Hsu

In the field of domain adaptation (DA) on 3D object detection, most of the work is dedicated to unsupervised domain adaptation (UDA). Yet, without any target annotations, the performance gap between the UDA approaches and the fully-supervised approach is still noticeable, which is impractical for real-world applications. On the other hand, weakly-supervised

Sidali Mohammdi, Matteo Bina, Abdelhakim Gharbi, Matteo G. A. Paris

We consider quantum systems with a Hamiltonian containing a weak perturbation i.e. $\boldsymbol{H=H_0} + \boldsymbol{\lambda} \cdot \boldsymbol{\tilde{H}}$, $\boldsymbol{\lambda}= \{\lambda_1, \lambda_2,...\}$, $\boldsymbol{\tilde{H}}$ $= \{H_1, H_2,...\}$, $\left|\boldsymbol{\lambda}\right| \ll 1$, and address situations where $\boldsymbol{\tilde{H}}$ is kn

Eric Beyerle, Pratyush Tiwary

Ligand unbinding is mediated by the free energy change, which has intertwined contributions from both energy and entropy. It is important but not easy to quantify their individual contributions. We model hydrophobic ligand unbinding for two systems, a methane particle and a C60 fullerene, both unbinding from hydrophobic pockets in all-atom water. By using a

P. M. Dee, J. S. Kim, A. C. Hire, J. Lim

Recent investigations into MoB$_{2}$ have unveiled a direct connection between a pressure-induced structural transition to a P6/mmm space group structure and the emergence of superconductivity, producing critical temperatures up to 32 K at 100 GPa. This pressure-induced superconducting state underscores the potential of doped MoB$_{2}$ as a possible candidat

Pablo Barcelo, Alexander Kozachinskiy, Anthony Widjaja Lin, Vladimir Podolskii

We contribute to the study of formal languages that can be recognized by transformer encoders. We focus on two self-attention mechanisms: (1) UHAT (Unique Hard Attention Transformers) and (2) AHAT (Average Hard Attention Transformers). UHAT encoders are known to recognize only languages inside the circuit complexity class ${\sf AC}^0$, i.e., accepted by a fa

Jarmo Hietarinta, Paul Martin, Eric C. Rowell

We find all solutions to the constant Yang--Baxter equation $R_{12}R_{13}R_{23}=R_{23}R_{13}R_{12}$ in three dimensions, subject to an additive charge-conservation ansatz. This ansatz is a generalisation of (strict) charge-conservation, for which a complete classification in all dimensions was recently obtained. Additive charge-conservation introduces additi

Paul Romatschke

Triviality of $\phi^4$ theory in four dimensions can be avoided if the bare coupling constant is negative in the UV. Theories with negative coupling can be put on the lattice if the integration domain for $\phi(x)$ is contour-deformed from the real to the complex domain. In 0+1d (quantum mechanics), one can recover results from $\mathcal{PT}$-symmetric quant

Zhong Guo, Aditya Chaudhari, Austin R. Coffman, Prabir Barooah

We consider the problem of optimal control of district cooling energy plants (DCEPs) consisting of multiple chillers, a cooling tower, and a thermal energy storage (TES), in the presence of time-varying electricity price. A straightforward application of model predictive control (MPC) requires solving a challenging mixed-integer nonlinear program (MINLP) bec

Fatemeh Farajzadeh, Ryan B. Killea, Alexander Teytelboym, Andrew C. Trapp

The United States has introduced a special humanitarian parole process for Ukrainian citizens in response to Russia 2022 invasion of Ukraine. To qualify for parole, Ukrainian applicants must have a sponsor in the United States. In collaboration with HIAS, a refugee resettlement agency involved in the parole process, we deployed RUTH (Refugees Uniting Through

Hyunsik Jeon, Jong-eun Lee, Jeongin Yun, U Kang

How can we recommend cold-start bundles to users? The cold-start problem in bundle recommendation is crucial because new bundles are continuously created on the Web for various marketing purposes. Despite its importance, existing methods for cold-start item recommendation are not readily applicable to bundles. They depend overly on historical information, ev

T. Lucas Makinen, Justin Alsing, Benjamin D. Wandelt

Set-based learning is an essential component of modern deep learning and network science. Graph Neural Networks (GNNs) and their edge-free counterparts Deepsets have proven remarkably useful on ragged and topologically challenging datasets. The key to learning informative embeddings for set members is a specified aggregation function, usually a sum, max, or

Costantino Pacilio, Davide Gerosa, Swetha Bhagwat

Combining multiple gravitational-wave observations allows for stringent tests of general relativity, targeting effects that would otherwise be undetectable using single-event analyses. We highlight how the finite size of the observed catalog induces a significant source of variance. If not appropriately accounted for, general relativity can be excluded with

Rafał Rechciński, Aleksei Khindanov, Dmitry I. Pikulin, Jian Liao

Topological insulator/superconductor two-dimensional heterostructures are promising candidates for realizing topological superconductivity and Majorana modes. In these systems, a vortex pinned by a pre-fabricated antidot in the superconductor can host Majorana zero-energy modes (MZMs), which are exotic quasiparticles that may enable quantum information proce

Iris Ulčakar, Zala Lenarčič

Integrable systems offer rare examples of solvable many-body problems in the quantum world. Due to the fine-tuned structure, their realization in nature and experiment is never completely accurate, therefore effects of integrability are observed only transiently. One way to overcome this limitation is to weakly couple nearly integrable systems to baths and d

Berkay Turan, Spencer Hutchinson, Mahnoosh Alizadeh

We introduce a novel algorithm for solving network utility maximization (NUM) problems that arise in resource allocation schemes over networks with known safety-critical constraints, where the constraints form an arbitrary convex and compact feasible set. Inspired by applications where customers' demand can only be affected through posted prices and real-tim

Sajad Abbar, Hiroki Nagakura

Neutrinos in dense environments like core-collapse supernovae (CCSNe) and neutron star mergers (NSMs) can undergo fast flavor conversions (FFCs) once the angular distribution of neutrino lepton number crosses zero along a certain direction. Recent advancements have demonstrated the effectiveness of machine learning (ML) in detecting these crossings. In this

Ka Ho Yuen, Hui Li, Huirong Yan

The temporal property of the compressible magneto-hydrodynamic (MHD) turbulence remains a fundamental unsolved question. Recent studies based on the spatial-temporal analysis in the global frame of reference suggest that the majority of fluctuation power in turbulence does not follow any of the MHD wave dispersion relations but has very low temporal frequenc

David Grant, Katherine Blundell, Emma Godden, Steven Lee

$\eta$ Carinae is an extremely luminous and energetic colliding-wind binary. The combination of its orbit and orientation, with respect to our line of sight, enables direct investigation of the conditions and geometry of the colliding winds. We analyse optical He I 5876 and 7065 $\unicode{x212B}$ line profiles from the Global Jet Watch observatories covering

Alex J. Vernon, Sebastian Golat, Claire Rigouzzo, Eugene A. Lim

Light carries intrinsic spin angular momentum (SAM) when the electric or magnetic field vector rotates over time. A familiar vector equation calculates the direction of light's SAM density using the right hand rule with reference to the electric and magnetic polarisation ellipses. Using Maxwell's equations, this vector equation can be decomposed into a sum o

Samuel Federman, S. Thomas Megeath, Adam E. Rubinstein, Robert Gutermuth

Investigating Protostellar Accretion is a Cycle 1 JWST program using the NIRSpec+MIRI integral field units to obtain 2.9--28 $\mu$m spectral cubes of five young protostars with luminosities of 0.2-10,000 L$_{\odot}$ in their primary accretion phase. This paper introduces the NIRSpec 2.9--5.3 $\mu$m data of the inner 840-9000 au with spatial resolutions from

Daniel A. Yahalomi, David Kipping, David Nesvorný, Paul A. Dalba

Transit Timing Variations (TTVs) can be induced by a range of physical phenomena, including planet-planet interactions, planet-moon interactions, and stellar activity. Recent work has shown that roughly half of moons would induce fast TTVs with a short period in the range of two-to-four orbits of its host planet around the star. An investigation of the Keple

Barak Rom, Re'em Sari, Dong Lai

Many stellar-mass Black Holes (sBHs) are expected to orbit supermassive black holes at galactic centers. For galaxies with Active Galactic Nuclei (AGN), it is likely that the sBHs reside in a disk. We study the formation of sBH binaries via gravitational wave emission in such disks. We examine analytically the dynamics of two sBHs orbiting a supermassive bla

F. Abbate, A. Ridolfi, P. C. C. Freire, P. V. Padmanabh

We present the results of observations aimed at discovering and studying pulsars in the core-collapsed globular cluster (GC) NGC 6522 performed by the MeerTIME and TRAPUM Large Survey Project with the MeerKAT telescope. We have discovered two new isolated pulsars bringing the total number of known pulsars in the cluster to six. PSR J1803$-$3002E is a mildly

Claire E. Williams, William Lake, Smadar Naoz, Blakesley Burkhart

The James Webb Space Telescope (JWST) is capable of probing extremely early eras of our Universe when the supersonic relative motions between dark matter and baryonic overdensities modulate structure formation ($z>\sim 10$). We study low-mass galaxy formation including this "stream velocity" using high resolution AREPO hydrodynamics simulations, and present

Majed Khalaf, Ofri Telem

Can classical systems be described analytically at all orders in their interaction strength? For periodic and approximately periodic systems, the answer is yes, as we show in this work. Our analytical approach, which we call the \textit{Quantum Spectral Method}, is based on a novel application of Bohr's correspondence principle, obtaining non-perturbative cl

Lena Janshen, Domenico Giulini

We investigate the asymptotic symmetry group of a SU(2)-Yang-Mills theory coupled to a Higgs field in the Hamiltonian formulation. This extends previous work on the asymptotic structure of pure electromagnetism by Henneaux and Troessaert, and on electromagnetism coupled to scalar fields and pure Yang-Mills fields by Tanzi and Giulini. We find that there are

Yunjing Wu, Zheng Cai, Jianan Li, Kristian Finlator

We report the first statistical analyses of [CII] and dust continuum observations in six strong OI absorber fields at the end of the reionization epoch obtained by the Atacama Large Millimeter/Submillimeter Array (ALMA). Combined with one [CII] emitter reported in Wu et al. (2021), we detect one OI-associated [CII] emitter in six fields. At redshifts of OI-a

Jean J. Somalwar, Vikram Ravi, Wenbin Lu

The multiwavelength properties of radio-emitting tidal disruption events (TDEs) are poorly understood. In a previous paper, we presented the first sample of radio-selected, optically-detected TDEs, which included two events (VT J1008 and VT J2012) associated with late-time (${\sim}2$ years post-optical flare) intermediate with emission lines that are largely

Yutaka Fujita, Takuma Izumi, Hiroshi Nagai, Nozomu Kawakatu

The physical quantity that directly controls the feedback of active galactic nuclei (AGNs) in elliptical galaxies remains to be determined. The discovery of molecular gas around the AGNs suggests that the gas is fueling the AGNs. Therefore, we analyze Atacama Large Millimeter/submillimeter Array (ALMA) data for the CO line (J=1-0, 2-1, 3-2) emission and esti

Naushad Ahmad Kamar, Adrian Kantian, Thierry Giamarchi

We study the dynamics of a mobile impurity in a two-leg bosonic ladder. The impurity moves both along and across the legs and interacts with a bath of interacting bosonic particles present in the ladder. We use both analytical (Tomonaga-Luttinger liquid - TLL) and numerical (Density Matrix Renormalization Group - DMRG) methods to compute the Green's function

Ward S. Howard, Adam F. Kowalski, Laura Flagg, Meredith A. MacGregor

We present the first analysis of JWST near-infrared spectroscopy of stellar flares from TRAPPIST-1 during transits of rocky exoplanets. Four flares were observed from 0.6--2.8 $\mu$m with NIRISS and 0.6--3.5 $\mu$m with NIRSpec during transits of TRAPPIST-1b, f, and g. We discover P$\alpha$ and Br$\beta$ line emission and characterize flare continuum at wave

Jean J. Somalwar, Vikram Ravi, Dillon Z. Dong, Erica Hammerstein

In this work, we use the Jansky VLA Sky Survey (VLASS) to compile the first sample of six radio-selected tidal disruption events (TDEs) with transient optical counterparts. While we still lack the statistics to do detailed population studies of radio-selected TDEs, we use these events to suggest trends in host galaxy and optical light curve properties that m

Thales A. Gutcke

The origin and formation of globular clusters has remained a mystery. We present a formation scenario for ancient globular cluster-like objects that form in ultra-high resolution simulations (smallest cell size $<0.1\,\mathrm{pc}$, mass resolution $M_\mathrm{cell}=4\,\mathrm{M}_\odot$). The simulations are cosmological zoom-in simulations of dwarf galaxies w

Noa Rubin, Inbar Seroussi, Zohar Ringel

A key property of deep neural networks (DNNs) is their ability to learn new features during training. This intriguing aspect of deep learning stands out most clearly in recently reported Grokking phenomena. While mainly reflected as a sudden increase in test accuracy, Grokking is also believed to be a beyond lazy-learning/Gaussian Process (GP) phenomenon inv

Yitian Sun, Katelin Schutz, Harper Sewalls, Calvin Leung

Axion dark matter (DM) produces echo images of bright radio sources via stimulated decay. These images appear as a faint radio line centered at half the axion mass, with the line width set by the DM velocity dispersion. Due to the kinematics of the decay, the echo can be emitted in the direction nearly opposite to the incoming source of stimulating radiation

R. Gottumukkala, L. Barrufet, P. A. Oesch, A. Weibel

With the advent of JWST, we can probe the rest-frame optical emission of galaxies at $z>3$ with high sensitivity and spatial resolution, making it possible to accurately characterise red, optically-faint galaxies and thus move towards a more complete census of the galaxy population at high redshifts. To this end, we present a sample of 148 massive, dusty gal

Lakshya Bhardwaj, Lea E. Bottini, Daniel Pajer, Sakura Schafer-Nameki

We propose a unified framework to classify gapped infra-red (IR) phases with categorical symmetries, leading to a generalized, categorical Landau paradigm. This is applicable in any dimension and gives a succinct, comprehensive, and computationally powerful approach to classifying gapped symmetric phases. The key tool is the symmetry topological field theory

Quentin Henry, François Larrouturou, Christophe Le Poncin-Lafitte

The aim of this work is twofold: (i) to properly define a wave-generation formalism for compact-supported sources embedded in Einstein-Maxwell theory, relying on matched post-Newtonian and multipolar-post-Minkowskian expansions; (ii) to apply this formalism (which is valid for any type of post-Newtonian sources) to the case of two stars with constant and ali

Lakshya Bhardwaj, Lea E. Bottini, Daniel Pajer, Sakura Schafer-Nameki

We propose a general framework to characterize gapped infra-red (IR) phases of theories with non-invertible (or categorical) symmetries. In this paper we focus on (1+1)d gapped phases with fusion category symmetries. The approach that we propose uses the Symmetry Topological Field Theory (SymTFT) as a key input: associated to a field theory in d spacetime di

Meredith Neyer, Aaron Smith, Rahul Kannan, Mark Vogelsberger

An important characteristic of cosmic hydrogen reionization is the growth of ionized gas bubbles surrounding early luminous objects. Ionized bubble sizes are beginning to be probed using Lyman-$\alpha$ emission from high-redshift galaxies, and will also be probed by upcoming 21-cm maps. We present results from a study of bubble sizes using the state-of-the-a

Jean J. Somalwar, Vikram Ravi, Yuhan Yao, Muryel Guolo

Tidal disruption events (TDEs) occur when a star enters the tidal radius of a supermassive black hole (SMBH). If the star only grazes the tidal radius, a fraction of the stellar mass will be accreted in a partial TDE (pTDE). The remainder can continue orbiting and may re-disrupted at pericenter, causing a repeating pTDE. pTDEs may be as or more common than f

Haotian Liu, Chunyuan Li, Yuheng Li, Yong Jae Lee

Large multimodal models (LMM) have recently shown encouraging progress with visual instruction tuning. In this note, we show that the fully-connected vision-language cross-modal connector in LLaVA is surprisingly powerful and data-efficient. With simple modifications to LLaVA, namely, using CLIP-ViT-L-336px with an MLP projection and adding academic-task-ori

Mengyu Yang, Patrick Grady, Samarth Brahmbhatt, Arun Balajee Vasudevan

How easy is it to sneak up on a robot? We examine whether we can detect people using only the incidental sounds they produce as they move, even when they try to be quiet. We collect a robotic dataset of high-quality 4-channel audio paired with 360 degree RGB data of people moving in different indoor settings. We train models that predict if there is a moving

Nils Blach, Maciej Besta, Daniele De Sensi, Jens Domke

Novel low-diameter network topologies such as Slim Fly (SF) offer significant cost and power advantages over the established Fat Tree, Clos, or Dragonfly. To spearhead the adoption of low-diameter networks, we design, implement, deploy, and evaluate the first real-world SF installation. We focus on deployment, management, and operational aspects of our test

Beatriz Tucci, Fabian Schmidt

Cosmological inferences typically rely on explicit expressions for the likelihood and covariance of the data vector, which normally consists of a set of summary statistics. However, in the case of nonlinear large-scale structure, exact expressions for either likelihood or covariance are unknown, and even approximate expressions can become very cumbersome, de

Shaowei Liu, Yang Zhou, Jimei Yang, Saurabh Gupta

This paper presents a novel object-centric contact representation ContactGen for hand-object interaction. The ContactGen comprises three components: a contact map indicates the contact location, a part map represents the contact hand part, and a direction map tells the contact direction within each part. Given an input object, we propose a conditional genera

Mihir Prabhudesai, Anirudh Goyal, Deepak Pathak, Katerina Fragkiadaki

Text-to-image diffusion models have recently emerged at the forefront of image generation, powered by very large-scale unsupervised or weakly supervised text-to-image training datasets. Due to their unsupervised training, controlling their behavior in downstream tasks, such as maximizing human-perceived image quality, image-text alignment, or ethical image g

Stefan Smeu, Elena Burceanu, Emanuela Haller, Andrei Liviu Nicolicioiu

Novelty detection seeks to identify samples deviating from a known distribution, yet data shifts in a multitude of ways, and only a few consist of relevant changes. Aligned with out-of-distribution generalization literature, we advocate for a formal distinction between task-relevant semantic or content changes and irrelevant style changes. This distinction f

Pok Man Tam, Charles L. Kane

A Fermi gas of non-interacting electrons, or ultra-cold fermionic atoms, has a quantum ground state defined by a region of occupancy in momentum space known as the Fermi sea. The Euler characteristic $\chi_F$ of the Fermi sea serves to topologically classify these gapless fermionic states. The topology of a $D$ dimensional Fermi sea is physically encoded in

Andrei Constantinescu, Roger Wattenhofer

An electorate with fully-ranked innate preferences casts approval votes over a finite set of alternatives. As a result, only partial information about the true preferences is revealed to the voting authorities. In an effort to understand the nature of the true preferences given only partial information, one might ask whether the unknown innate preferences co

Yi-Kai Liu

The curvelet transform is a special type of wavelet transform, which is useful for estimating the locations and orientations of waves propagating in Euclidean space. We prove an uncertainty principle that lower-bounds the variance of these estimates, for radial wave functions in n dimensions. As an application of this uncertainty principle, we show the infea

Tianhong Li, Sangnie Bhardwaj, Yonglong Tian, Han Zhang

Current vision-language generative models rely on expansive corpora of paired image-text data to attain optimal performance and generalization capabilities. However, automatically collecting such data (e.g. via large-scale web scraping) leads to low quality and poor image-text correlation, while human annotation is more accurate but requires significant manu

Ryan C. Challener, Luis Welbanks, Peter McGill

Eclipse mapping is a technique for inferring 2D brightness maps of transiting exoplanets from the shape of an eclipse light curve. With JWST's unmatched precision, eclipse mapping is now possible for a large number of exoplanets. However, eclipse mapping has only been applied to two planets and the nuances of fitting eclipse maps are not yet fully understood

Trevor Arp, Owen Sheekey, Haoxin Zhou, C. L. Tschirhart

Rhombohedral graphene multilayers provide a clean and highly reproducible platform to explore the emergence of superconductivity and magnetism in a strongly interacting electron system. Here, we use electronic compressibility and local magnetometry to explore the phase diagram of this material class in unprecedented detail. We focus on rhombohedral trilayer

E. N. Antonov, A. Yu. Orlov, D. V. Vasiliev

Explicit expressions for multimatrix models with complex and unitary matrices allows to couple these models with well-known unitary, orthogonsl and sympletic ensembles. We consider examples of such mixed ensembles which are solvable in the sense that the partition functions of such ensembles can be considered as tau functions of the classical integrable equa

Ke Wang, Houxing Ren, Aojun Zhou, Zimu Lu

The recently released GPT-4 Code Interpreter has demonstrated remarkable proficiency in solving challenging math problems, primarily attributed to its ability to seamlessly reason with natural language, generate code, execute code, and continue reasoning based on the execution output. In this paper, we present a method to fine-tune open-source language model

Sachidananda Barik, Akhil Mohanan, Sayantan Majumdar

Increase in viscosity under increasing shear stress, known as shear thickening (ST), is one of the most striking properties of dense particulate suspensions. Under appropriate conditions, they exhibit discontinuous shear thickening (DST), where the viscosity increases dramatically and can also transform into a solid-like state due to shear induced jamming (S

Sungguen Ryu, Rosa López, Llorenç Serra, David Sanchez

Identifying and controlling decoherence in single electron sources (SES) is important for their applications in quantum information processing. The recent experiments with ultrashort electron pulses [J. D. Fletcher et al., Nat. Commun. 10, 5298 (2019)] demonstrate strong decoherence that cannot be caused by traditional mechanisms such as electron-electron or

Renan B. Magalhães, Andreu Masó-Ferrando, Flavio Bombacigno, Gonzalo J. Olmo

We construct a general class of modified Ellis wormholes, where one asymptotic Minkowski region is replaced by a bounded 2-sphere core, characterized by asymptotic finite areal radius. We pursue an in-depth analysis of the resulting geometry, outlining that geodesic completeness is guaranteed also when the radial function asymptotically shrinks to zero. Then

Marin Ðujić, D. Buhin, N. Šantić, D. Aumiler

We perform a comparative study of light storage in antirelaxation-coated and buffer-gas-filled alkali-vapor cells using electromagnetically induced transparency (EIT) in warm rubidium vapor. The use of a buffer-gas-filled cell resulted in $\approx$10-fold improvement in storage time and efficiency compared to antirelaxation coated cells. Our findings contrib

Michael S. Albergo, Mark Goldstein, Nicholas M. Boffi, Rajesh Ranganath

Generative models inspired by dynamical transport of measure -- such as flows and diffusions -- construct a continuous-time map between two probability densities. Conventionally, one of these is the target density, only accessible through samples, while the other is taken as a simple base density that is data-agnostic. In this work, using the framework of st

Paul-Ambroise Duquenne, Holger Schwenk, Benoît Sagot

Recent research has shown that independently trained encoders and decoders, combined through a shared fixed-size representation, can achieve competitive performance in speech-to-text translation. In this work, we show that this type of approach can be further improved with multilingual training. We observe significant improvements in zero-shot cross-modal sp

Viktor Christiansson, Philipp Werner

Using ab-initio density functional theory for superconductors (SCDFT), we systematically study the quaternary borocarbides $RM_2$B$_2$C. Treating the retarded (frequency-dependent) interaction $W(\omega)$ within the random-phase approximation (RPA), we find good agreement with experiments for the calculated superconducting critical temperature $T_c$ in the n

Hongjian Wang, Aaditya Ramdas

In 1976, Lai constructed a nontrivial confidence sequence for the mean $\mu$ of a Gaussian distribution with unknown variance $\sigma^2$. Curiously, he employed both an improper (right Haar) mixture over $\sigma$ and an improper (flat) mixture over $\mu$. Here, we elaborate carefully on the details of his construction, which use generalized nonintegrable mar

Debasish Borah, Satyabrata Mahapatra, Dibyendu Nanda, Sujit Kumar Sahoo

We study the possibility of generating light Dirac neutrino mass via scotogenic mechanism where singlet-doublet fermion dark matter (DM) plays non-trivial role in generating one-loop neutrino mass, anomalous magnetic moment of muon $(g-2)_\mu$ as well as additional relativistic degrees of freedom $\Delta{N_{\rm eff}}$ within reach of cosmic microwave backgro

Paloma Sodhi, S. R. K. Branavan, Yoav Artzi, Ryan McDonald

Performing tasks on the web presents fundamental challenges to large language models (LLMs), including combinatorially large open-world tasks and variations across web interfaces. Simply specifying a large prompt to handle all possible behaviors and states is extremely complex, and results in behavior leaks between unrelated behaviors. Decomposition to disti

Michael Werner, Philip K. Schwartz, Jan-Niclas Kirsten-Siemß, Naceur Gaaloul

We present a systematic approach to determine all relativistic phases up to $\mathcal{O}(c^{-2})$ in light-pulse atom interferometers in weakly curved spacetime that are based on elastic scattering, namely Bragg diffraction and Bloch oscillations. Our analysis is derived from first principles using the parameterized post-Newtonian formalism. In the treatment

Yihang Yao, Zuxin Liu, Zhepeng Cen, Jiacheng Zhu

Safe reinforcement learning (RL) focuses on training reward-maximizing agents subject to pre-defined safety constraints. Yet, learning versatile safe policies that can adapt to varying safety constraint requirements during deployment without retraining remains a largely unexplored and challenging area. In this work, we formulate the versatile safe RL problem

Prasann Singhal, Tanya Goyal, Jiacheng Xu, Greg Durrett

Great success has been reported using Reinforcement Learning from Human Feedback (RLHF) to align large language models, with open preference datasets enabling wider experimentation, particularly for "helpfulness" in tasks like dialogue and web question answering. Alongside these improvements, however, RLHF also often drives models to produce longer outputs.

Nestor Maslej, Loredana Fattorini, Erik Brynjolfsson, John Etchemendy

Welcome to the sixth edition of the AI Index Report. This year, the report introduces more original data than any previous edition, including a new chapter on AI public opinion, a more thorough technical performance chapter, original analysis about large language and multimodal models, detailed trends in global AI legislation records, a study of the environm

Omar Khattab, Arnav Singhvi, Paridhi Maheshwari, Zhiyuan Zhang

The ML community is rapidly exploring techniques for prompting language models (LMs) and for stacking them into pipelines that solve complex tasks. Unfortunately, existing LM pipelines are typically implemented using hard-coded "prompt templates", i.e. lengthy strings discovered via trial and error. Toward a more systematic approach for developing and optimi

Matthew C. Nixon, Luis Welbanks, Peter McGill, Eliza M. -R. Kempton

A key goal of exoplanet spectroscopy is to measure atmospheric properties, such as abundances of chemical species, in order to connect them to our understanding of atmospheric physics and planet formation. In this new era of high-quality JWST data, it is paramount that these measurement methods are robust. When comparing atmospheric models to observations, m

Anderson D. S. Duraes, J. Daniel Gezelter

We present a theory for pitch, a matrix property which is linked to the coupling of rotational and translational motion of rigid bodies at low Reynolds number. The pitch matrix is a geometric property of objects in contact with a surrounding fluid, and it can be decomposed into three principal axes of pitch and their associated \textit{moments of pitch}. The

Qiu Sujuan, Hoernisa Iminniyaz

Based on the cosmological observations, the baryon and dark matter energy densities are similar. There may exist an asymmetry between the dark matter particles and their corresponding anti--particles. We review the relic density of asymmetric dark matter particle and anti--particle in the standard cosmological scenario and quintessence model with kination ph

Nicholas Crispino, Kyle Montgomery, Fankun Zeng, Dawn Song

We introduce a method to improve the zero-shot reasoning abilities of large language models on general language understanding tasks. Specifically, we build an autonomous agent to instruct the reasoning process of large language models. We show this approach further unleashes the zero-shot reasoning abilities of large language models to more tasks. We study t

Irmak Balçık, Burton Newman

Let $K$ be a quadratic number field and let $E$ be an elliptic curve defined over $K$ such that $E[2] \not\subseteq E(K).$ In this paper, we study the effect of quadratic base change on $E(K)_{\text{tor}}.$ Moreover, for a given elliptic curve $E/K$ with prescribed torsion group over $K,$ (no restriction on its $2$-torsion part) we describe a fast algorithm

Zhanhui Zhou, Jie Liu, Jing Shao, Xiangyu Yue

A single language model, even when aligned with labelers through reinforcement learning from human feedback (RLHF), may not suit all human preferences. Recent approaches therefore prefer customization, gathering multi-dimensional feedback, and creating distinct reward models for each dimension. Different language models are then optimized for various prefere

Ofir Bar Tal, Adi Haviv, Amit H. Bermano

Evasion Attacks (EA) are used to test the robustness of trained neural networks by distorting input data to misguide the model into incorrect classifications. Creating these attacks is a challenging task, especially with the ever-increasing complexity of models and datasets. In this work, we introduce a self-supervised, computationally economical method for